Little additional risk of Gorkha earthquake-induced outburst floods found for many Himalayan lakes 

Note: This report is an update of two prior reports:

Much of Nepal and bordering areas of the Tibet Autonomous Region of China were heavily shaken by the magnitude 7.8 earthquake on 25 April 2015 and its aftershock of magnitude 7.3 (USGS). After the two major tremors, scientists and authorities in Nepal, from the U.S. Geological Survey (USGS), and from a NASA- and University of Arizona-led international volunteer-group of satellite image analysts have been examining the condition of Nepal’s most dangerous glacial lakes. Over 300 lakes have been examined. This report provides an update of the situation at Tsho Rolpa and Imja Tsho, and the potential for a glacial lake outburst flood (GLOF) from them. Tsho Rolpa has been especially worrisome due to its location near the epicentre of the largest aftershock.

Study of a satellite image of Tsho Rolpa taken on 17 May – five days after the nearby magnitude 7.3 aftershock – by NASA’s EO-1 satellite, and more recently by the Japan/U.S. instrument ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) aboard the NASA Terra satellite, shows no definitive evidence that Tsho Rolpa’s moraine or other parts of the glacier or lake had been damaged. The 10 m (33 ft) resolution of the EO-1 image and 15 m resolution ASTER image do not permit detailed assessment of possible structural features smaller than  their resolution. However, what can be seen does not elicit new concerns about the potential for a GLOF, and these new observations erase some of the worst fears that heavy damage to this part of the landscape had been incurred. For example, no obvious new drainage ways from Tsho Rolpa and no large landslides into the lake can be easily discerned (Figures 1 and 2). Discernible differences in pre- and post-earthquake images are primarily related to changes in seasonal snow cover. 

Photographs taken from a helicopter on 21 May 2015 and released by a Nepali newspaper ( show no visible disturbance to the moraine dam or the engineered structures used to lower and control the lake level. Satellite images and these aerial photos show that the lake is well-contained within the moraine dam at this time. This is further corroborated by a photo taken 27 May 2015 during a helicopter mission organized by the USGS (Figure 3), in which the moraine is intact as far as can be seen. Additional aerial and ground-based photos were obtained by an independent concerned citizen, Saroj Dhoj Joshi, and were made available for this article with his permission ( These photos show the hazardously poised lateral moraine and adjacent mountain slopes, yet they also show no indication of recent disturbances caused by the earthquakes. This work is an example of how satellite imaging, expert knowledge, and ’citizen science’ can work together to contribute science to discussions of public safety in the aftermath of a natural disaster. 

However, good news is always tentative when mountain hazards are concerned. Though the moraine remains intact (Figure 3), cracks in the moraine – not necessarily of a type that would likely cause moraine failure – were also observed during the USGS helicopter mission (Figure 4). According to reports from Nepal’s Department of Hydrology and Meteorology, unspecified “cracks” might have formed after the 12 May aftershock. These reported cracks are presumably the same as those visible in Figure 4; however, the actual date of their formation is not known. Because this part of the moraine does not contain ice, it is thought that the cracks are not likely to induce a failure of the moraine dam. Instead, the cracks may amount to a relatively mild slumping of some moraine material toward the lake. However, there is still a lack of information about how the deeper structure of the moraine may or may not have been affected by the earthquake.

There have been long-running concerns about Tsho Rolpa’s safety. A hazard mitigation project 15 years ago successfully lowered the lake level, which has helped reduce the possible impacts in the event of a GLOF. A GLOF early warning system based on meteoburst technology was installed during the project, but this system was rendered dysfunctional as a result of vandalism and a lack of maintenance during the insurgency period, mainly between 2000 and 2005. As authorities recognized the importance of the issue and residents’ heightened anxieties after the Gorkha earthquake, this month a new automatic GLOF-warning system was installed ( The new system will alert authorities, who can alert residents of downstream villages such as Na in the event of a GLOF or a precursor event.

Further welcome news from several partners in the volunteer group is that many other glacial lakes in the Himalayas are also seemingly untouched by the earthquake; however, this analysis is limited to the visibility of critical features in the satellite images used. Satellite images from lakes in the Tibet Autonomous Region of China, just across the border from Tsho Rolpa, show no evidence of outburst floods or major damage to their moraines (Figure 5). The Chinese Academy of Sciences (CAS) is tracking the effects of the earthquake in Tibet. Although many landslides have occurred there, the moraine dams of glacier lakes that have been checked by the CAS team appear to be unaffected by the quake. Nagoya University (Japan) and the University of Dayton (Ohio, USA) have already conducted independent satellite surveys of most of the hazardous glacial lakes in Nepal. So far the surveys have indicated landslides or avalanches reaching only eight of the more than 400 surveyed lakes, but none has emitted a flood as far as can be detected. Satellite images of Thulagi and Imja lakes have been assessed, and the two lakes appear free from visible damage. With one possible minor exception, no media reports from either Nepal or China have indicated the occurrence of a GLOF due to seismic activity in 2015. 

One minor exception is a report widely circulated on social media that a small GLOF may have occurred on 24 May in the vicinity of Imja Lake, as reflected in residents’ reporting of a loud sound heard for several kilometres – possibly from Lhotse Glacier (south of Everest) – and then a transient, anomalous rise in stream level. However, there is no evidence that points toward Imja Lake being the source. Instead, this event has been interpreted as ice fracturing and a small supraglacial pond draining from Lhotse Glacier. This would be the only GLOF reported from the region in the weeks following the Gorkha earthquake. 

The situation of Himalayan glacial lakes must be watched carefully in China, India, and Nepal to confirm our preliminary conclusions that the GLOF risk has not been visibly heightened by the earthquake. There are many glacial lakes in the Himalayas, many of which are in remote areas that are hard to reach. Because of this, satellite imagery will have to be used to inspect a large portion of the region’s glacial lakes. However, detailed assessments using aerial surveys and ground-based observations are needed to verify that no damage was done on scales smaller than what is visible in satellite images, particularly for select priority glacial lakes. 

As with the entire area shaken by the earthquake, there are concerns about how the coming monsoon may affect small-scale fractures or other damage not revealed by satellite images. Although many of these lakes remain a danger – Rolpa, Imja, and Thulagi (Dona) among them – our findings have not uncovered satellite-based evidence that their condition was changed by the earthquake. However, the limitations of using the available satellite imagery are such that it is not possible to see some details, and some features may be missed. However, the observations to date are encouraging on the whole. 

Contributors: Jeffrey S. Kargel, University of Arizona, Tucson, AZ, USA, ; Koji Fujita, Nagoya University, ; Umesh Haritashya, University of Dayton, ; Saroj Dhoj Joshi, Real Time Solutions Pvt Ltd, Lalitpur, Nepal; Sharad Joshi, International Centre for Integrated Mountain Development (ICIMOD), Kathmandu, ; Narendra Raj Khanal, ICIMOD, Kathmandu, ; Damodar Lamsal, Nagoya University (); Gregory Leonard, University of Arizona, Tucson, AZ, ; Akiko Seppyo Sakai, Nagoya University, ; Liu Shiyin, Cold and Arid Regions Environmental and Engineering Research Institute (CAREERI), Chinese Academy of Sciences (CAS), Lanzhou, Gansu, China, ; Arun Shrestha, ICIMOD, Kathmandu, ; Dan Shugar, University of Victoria, .

Figure 1: Tsho Rolpa relative to Na village. Na, just 3,600 m (2.2 miles) downstream from Tsho Rolpa, is the first major village at serious risk of a potential GLOF from Tsho Rolpa.

Figure 2: Post-earthquake image of Tsho Rolpa appears basically identical to how it appeared shortly before the earthquake

Figure 3: The end moraine and lateral moraines containing Tsho Rolpa appear to be in the same state as before the earthquake, according to the USGS helicopter-borne inspection on 27 May 2015. Photo credit: Brian Collins/USGS-USAID-OFDA

Figure 4: Two areas of fractures (outlined in red) that were possibly, but not necessarily, formed by the 12 May aftershock were observed on the engineered part of the end moraine during a USGS inspection flight at Tsho Rolpa on 27 May. The photo also shows the drainage sluice and other infrastructure established in 2000 to lower and control the lake level. Photo credit: Brian Collins/USGS-USAID-OFDA

Figure 5: Lakes in the Tibet Autonomous Region of China just across the border from Tsho Rolpa. The red dots identify lakes that appear completely unchanged from their state just prior to the earthquake